Voice or speech quality is typically one of the major key performance indicators for telephones. However communication networks may set communication parameters for throughput optimization which are not optimal for real-time communication services such as speech data transmission. For example, while a relatively high BLER (block error rate) of e.g. 10% may be optimal in terms of throughput, voice quality may suffer from this high BLER, e.g. due to the caused delay of data packets. Accordingly, approaches to increase voice or speech quality, or generally the quality of communication services for which factors besides throughput, e.g. a low BLER, are of importance, are desirable.